Multi-hop Question Answering on Bamboogle (standard val)
40Exact Match (EM)StepSearch-instruct
Evaluation Results
| Method | Links | |
|---|---|---|
| StepSearch-instructType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 40 | |
| Apex-SearcherType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 40 | |
| DecEx-RAGType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 37.6 | |
| Search-R1-InstructType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 36.8 | |
| ReasonRAGType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 36 | |
| Apex-SearcherType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.03 | 35.2 | |
| StepSearch-instructType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.03 | 32.8 | |
| Search-o1Type=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.03 | 32 | |
| Search-o1Type=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 29.6 | |
| ZeroSearch-instructType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 27.8 | |
| IRCoTType=Iterative RAG, Backbone=Qwen2.5-3B-Instruct2026.03 | 24 | |
| CoTType=Non-Retrieval, Backbone=Qwen2.5-7B-Instruct2026.03 | 23.2 | |
| IRCoTType=Iterative RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 22.4 | |
| RAGType=Standard RAG, Backbone=Qwen2.5-7B-Instruct2026.03 | 20.8 | |
| Direct InferenceType=Non-Retrieval, Backbone=Qwen2.5-7B-Instruct2026.03 | 12 | |
| ZeroSearch-instructType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.03 | 11.1 | |
| Search-R1-InstructType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.03 | 8.8 | |
| RAGType=Standard RAG, Backbone=Qwen2.5-3B-Instruct2026.03 | 8 | |
| Direct InferenceType=Non-Retrieval, Backbone=Qwen2.5-3B-Instruct2026.03 | 2.4 | |
| CoTType=Non-Retrieval, Backbone=Qwen2.5-3B-Instruct2026.03 | 0 |